JP2001251515A - Image processor - Google Patents

Abstract

PROBLEM TO BE SOLVED: To equally improve image quality of a character part, a dot part, and characters on dots performing adaptive processing corresponding the feature quantity of an image, without using image area discriminating means for image area separation or the like. SOLUTION: A first filter means 6 subjects an input image to band emphasis, and a second filter means 7 extracts a high-band frequency component from the input image. An edge quantity is detected (8) from the input picture by a means, and the detected edge quantity is nonlinearly converted (9). The high- band component is edge enhanced (10) on the basis of the edge quantity after conversion and is added (11) to the input image 11. If the edge quantity is equal to or larger than a prescribed threshold, a selection means 12 selects an output signal from the adder 11; but otherwise the means 12 selects an output signal from the first filter means 6.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an image processing apparatus for improving the image quality of an output image in digital image processing, and more particularly to various types of originals such as a character original, a halftone original, and a character original on a halftone dot. This is a technique applied to, for example, a digital copying machine, a facsimile, and the like.

[0002]

2. Description of the Related Art In a conventional digital copying machine, adaptive processing is performed based on local information (for example, edge information) of a document in which characters and pictures are mixed, so that optimum processing can be performed on image data. Is going.

For example, by controlling the mixing ratio of the signal after the smoothing process and the signal after the edge emphasizing process in accordance with the output of the edge detecting means of the image signal, the character portion where the edge amount is largely detected is controlled. For example, an apparatus has been proposed in which an edge emphasis process is performed and a halftone portion having a small edge amount is subjected to a smoothing process (see Japanese Patent Application Laid-Open No. 61-157162).

However, according to the above-described edge detection method, the edge amount is large in the vicinity of the halftone dot character, so that the portion is less smoothed and the edge is strongly emphasized. In other words, the halftone dots in the vicinity of the character cannot be smoothed with an appropriate strength, and the area around the character becomes noisy. That is, halftone dots around the character remain due to weak smoothing, and as a result, a notch is attached to the character edge.

[0005]

An apparatus for solving the above-mentioned problems has been proposed by the present applicant (see Japanese Patent Application Laid-Open No. 7-95409). The apparatus has an image area separating circuit for separating a character area from a non-character area, and performs an optimal process based on the separation information.

Then, a method of improving reproducibility by processing characters on halftone dots for non-characters is adopted. That is, the input image signal is subjected to a smoothing process to attenuate a frequency component equal to or higher than a specific frequency, and the smoothed image signal is subjected to an adaptive edge enhancement process based on the edge amount. As a result, it is possible to enhance the edge of a character on a halftone dot while preventing the occurrence of moire at the halftone dot portion.

[0007] However, the above-mentioned technique can provide almost good sharpness for characters on halftone dots, but lacks sharpness when applied to characters on a white background. Further, in the above-described technique, although the image area separation circuit is provided, the cost is increased because the size of the image area separation circuit is large.

SUMMARY OF THE INVENTION The present invention has been made in view of the above-mentioned problems, and it is an object of the present invention to perform adaptive processing in accordance with the feature amount of an image without using image area discriminating means such as image area separation. Accordingly, it is an object of the present invention to provide an image processing apparatus which achieves both image quality of a character portion, a halftone dot portion, and a character on a halftone dot.

[0009]

According to the present invention, a character document,
Various kinds of originals such as halftone originals and halftone originals are subjected to suitable filtering. Further, in a halftone image portion having a small edge amount or a continuous tone image portion such as a photographic paper photograph, it is preferable to perform edge enhancement such that moire does not appear rather than to perform a smoothing process. Can be performed. According to the present invention, high image quality is realized even in an image area having a small edge amount.

In the present invention, high image quality is achieved without increasing the amount of hardware such as a line memory.

Further, in the present invention, the filter coefficient for an image with a large edge amount and the filter coefficient for an image with a small edge amount are partially multiplied by n times (n is an integer other than 0). By making them related, the processing is speeded up.

[0012]

DESCRIPTION OF THE PREFERRED EMBODIMENTS One embodiment of the present invention will be specifically described below with reference to the drawings. FIG. 2 is a diagram illustrating a configuration of a general image forming apparatus. An image signal of a document or the like input from the image input unit 1 is input to the density conversion unit 3 after the spatial frequency characteristics are corrected by the filter processing unit 2. The density conversion means 3 converts the gradation characteristics between the original and the image output by the image output means 5 so as to have predetermined gradation characteristics. Next, the halftone processing means 4 performs pseudo halftone processing, and the image output means 5 performs image output on paper or the like. The above-described image signal is a digital image signal having a value of 0 to 255 quantized by, for example, 8 bits.

The present invention relates to the filter processing means 2 in the above configuration. FIG. 1 shows a configuration of a filter processing unit according to an embodiment of the present invention. As shown in FIG. 1, an input image signal is input to a first filter processing unit 6, where a convolution operation is performed on a pixel of interest by a predetermined filter coefficient, and a frequency characteristic is corrected. The filter processing applied here is, for example, a coefficient as shown in FIG. 3 (the center position is the pixel of interest), which is a band for emphasizing a medium frequency band as shown in FIG. It has an enhanced frequency transfer characteristic. Thereby, it is configured such that an appropriate edge enhancement process can be performed while suppressing moire in a halftone image having 133 lines or more.

The input image signal is supplied to the edge amount detecting means 8.
And the edge amount E of the pixel of interest included in the image signal is detected. Here, the edge amount detecting means is, for example,
It is composed of a primary differential filter (the center position is the pixel of interest) as shown in FIGS. FIG. 5 detects a vertical edge component in an image, FIG. 6 detects a horizontal edge component, and FIGS. 7 and 8 detect an oblique edge component, respectively.

The edge amount detecting means 8 is configured to output the maximum value of the absolute value as the edge amount E among the edge detection amounts by these four primary differential filters. The edge amount E of the target pixel is output as the edge amount C converted by the edge amount conversion means 9. The edge amount conversion means 9 is, for example, a conversion table as shown in FIG. 9, and has a non-linear conversion characteristic such that the converted edge amount C is saturated for an edge amount E equal to or more than a predetermined amount. .

A second filter processing means 7 is provided separately from the first filter processing means. The second filter processing means 7 uses, for example, a secondary differential filter as shown in FIG. A predetermined spatial frequency component of the signal is extracted. The secondary differential filter (the center position is the pixel of interest) shown in FIG. 4 has a frequency transfer characteristic as shown in FIG. 11 and is a filter for extracting a high-frequency spatial frequency component.

Further, the multiplier 10 multiplies the frequency component extracted by the second filter processing means 7 by the signal C from the edge amount conversion means 9 and outputs the result to the adder 11. The adder 11 is configured to perform addition with the original input image signal and output the obtained signal B to the selection unit 12.

The selecting means 12 selectively outputs according to the signal C from the edge amount converting means 9. The selection means 12 has a predetermined threshold, and when the edge amount C is equal to or more than the predetermined threshold, the signal B from the second filter processing means 7
And outputs a signal A from the first filter processing means 6 when the value is less than a predetermined threshold value.

In this embodiment, an example of the primary differential filter of 5 × 5 size has been described as the edge amount detecting means 8, but the present invention is not limited to this. A filter may be used. In addition, almost the same effect can be obtained even if the edge detection in the oblique direction is not provided in order to reduce the hardware scale. Further, in the present embodiment, the example has been described in which the edge amount conversion means 9 non-linearly converts the edge amount E using a conversion table, but this may be realized by a conversion formula or the like.

Further, in this embodiment, an example has been shown in which the selecting means 12 selectively outputs the signal A or the signal B in accordance with the output signal C from the edge amount converting means 9.
The mixing ratio of the signal A and the signal B may be controlled based on the output signal C.

With the above configuration, an image region having a small edge amount such as a halftone dot portion having a high number of lines is processed by the first filter processing means having the characteristic of the band emphasis filter. For this reason, it is possible to correct the frequency characteristic in which the sharpness of the picture portion is improved while suppressing the moiré. Further, the edge portion processed by the second filter processing means 8 is subjected to adaptive edge enhancement processing based on the edge amount detected by the edge amount detection means 8 and the edge amount conversion means 9, so that the character portion An image with improved sharpness can be reproduced.

In the mixing of a smoothing filter and an emphasizing filter as described in the above-mentioned Japanese Patent Application Laid-Open No. 61-157162, the frequency transfer characteristics of the two filters are greatly different. Such a switching portion becomes unnatural, causing image quality deterioration.

According to the present invention, since a band emphasis filter and an adaptive edge emphasis filter are mixed, each of them has an emphasis characteristic in a medium frequency band.
No image degradation occurs at the switching portion as described above. In the technique disclosed in Japanese Patent Application Laid-Open No. 7-95409, adaptive edge enhancement is performed after a smoothing filter, and image quality is not degraded at a boundary portion, but enhancement of a character portion is insufficient. In contrast, in the present invention,
Sufficient edge emphasis is also performed on the character portion, and the image quality of the character portion, the halftone dot portion, and the character on the halftone dot can be compatible.

In the above embodiment, the first filter processing means 6 is an example of a band emphasizing filter. However, in the case where the occurrence of moiré in a halftone dot portion is further suppressed or the halftone dot shape is smoothed. For example, a smoothing filter as shown in FIG. 12 may be used. The frequency transfer characteristic of the smoothing filter shown in FIG. 12 has such a characteristic that the frequency characteristic is attenuated in a middle to high frequency region as shown in FIG.

When the first filter processing means 6 is a band emphasis filter, to realize a suitable edge emphasis while suppressing moiré, a 5 × 5 size or 7 × 7 filter is used.
A size coefficient matrix is required. For example, in the case of a 7 × 7 size matrix, when the embodiment of FIG. 1 is realized by hardware, six line memories are required. On the other hand, if the filter size of the second filter processing means 7 is larger than the size of the first filter processing means 6,
Further, a line memory is required.

Therefore, by setting the second filter size to be at least equal to or smaller than the first filter size, an increase in line memory can be prevented and the cost can be reduced.

Further, when the embodiment shown in FIG. 1 is realized by software, for example, the product-sum operation processing for performing the filter processing increases, and the processing becomes slow. Therefore,
In order to solve this, it is assumed that the filter coefficients of the first filter processing means 6 are as shown in FIG. 14, and the filter coefficients of the second filter processing means 7 are as shown in FIG. The coefficient group indicated by oblique lines in FIG. 15 is a value obtained by multiplying the coefficient group illustrated in FIG. 14 by (−1). In other words, the product-sum calculation result of the hatched portion obtained by the first filter processing unit 6 can be multiplied by (−1) and used as it is in the second filter processing unit 7.

As described above, a part of the filter coefficient of the second filter processing means is a coefficient obtained by multiplying a part of the filter coefficient of the first filter processing means by n (n is an integer other than 0). Thus, it is possible to realize a high-speed processing.

[0029]

As described above, according to the first and second aspects of the present invention, processing is performed using the band emphasizing filter and the adaptive edge emphasizing filter. The sharpness of the picture part can be improved, and sufficient edge emphasis is also performed on the character part,
The sharpness can be improved, and thereby the image quality of the character portion, the halftone dot portion, and the character on the halftone dot can be compatible.

According to the third aspect of the present invention, since the first filter processing means corrects the frequency characteristic so as to attenuate the frequency characteristic in the middle to high frequency region, the moire in the halftone dot portion can be further improved. Can be suppressed.

According to the fourth and fifth aspects of the present invention, the size of the second filter for extracting high-frequency components is configured to be smaller than the size of the first filter, thereby increasing the circuit scale. Is prevented, which can contribute to cost reduction.

According to the sixth aspect of the present invention, a part of the filter coefficients of the second filter processing means is n times (n is an integer other than 0) a part of the filter coefficients of the first filter processing means. Since the coefficients are configured so as to obtain the calculated results, it is possible to obtain a calculation result of the second filter processing means by performing a simple calculation processing on the calculation result of the first filter processing means, and to speed up the processing. it can.

[Brief description of the drawings]

FIG. 1 shows a configuration of a filter processing unit according to an embodiment of the present invention.

FIG. 2 shows a configuration of a general image forming apparatus.

FIG. 3 shows an example of a filter applied to a first filter processing means.

FIG. 4 shows an example of a filter applied to a second filter processing means.

FIG. 5 shows a first-order differential filter for detecting a vertical edge component.

FIG. 6 shows a primary differential filter for detecting a horizontal edge component.

FIG. 7 shows a first-order differential filter for detecting an edge component in an oblique direction.

FIG. 8 shows a first-order differential filter for detecting an edge component in a diagonal direction.

FIG. 9 shows conversion characteristics of an edge amount conversion table.

FIG. 10 shows characteristics of the filter of FIG. 3;

FIG. 11 shows characteristics of the filter of FIG.

FIG. 12 shows an example of a smoothing filter.

FIG. 13 shows characteristics of the filter of FIG.

FIG. 14 shows filter coefficients applied to the first filter processing means.

15 shows a filter coefficient (hatched portion) obtained by multiplying the coefficient of the hatched portion in FIG. 14 by (−1).

[Explanation of symbols]

 REFERENCE SIGNS LIST 1 image input means 2 filter processing means 3 density conversion means 4 halftone processing means 5 image output means 6 first filter processing means 7 second filter processing means 8 edge amount detection means 9 edge amount conversion means 10 multiplier 11 addition Vessel 12 selection means

 ──────────────────────────────────────────────────続 き Continued on the front page F term (reference) 5B057 AA11 BA02 BA30 CA02 CA06 CA07 CA08 CA12 CA16 CB02 CB06 CB07 CB08 CB12 CB16 CC02 CE03 CE05 CE06 CH07 CH09 CH18 DA17 DB02 DB05 DB08 DB09 DC16 5C077 LL17 LL18 LL19 MP02 MP03 MP05 PP09 PP27 PP28 PP47 PP49 PP61 PP68 PQ08 PQ12 PQ23

Claims (6)

[Claims] 1. An edge amount detecting means for detecting an edge amount of an input image signal, a first filter processing means for performing a first frequency characteristic conversion on the image signal, Second filter processing means for converting a second frequency characteristic, first calculation means for calculating the detected edge amount and output of the second filter processing means, and first calculation means Means for calculating the output of the means and the image signal, and the output of the first filter processing means or the second processing means in accordance with the detected edge amount.
Selecting means for selecting the output of the calculating means. 2. The image processing apparatus according to claim 1, wherein the first filter processing unit is a filter that emphasizes a spatial frequency component in a predetermined band. 3. The image processing apparatus according to claim 1, wherein said first filter processing means is a filter for attenuating a predetermined spatial frequency component. 4. An image processing apparatus according to claim 1, wherein said second filter processing means is a filter for extracting a predetermined spatial frequency component. 5. The image processing apparatus according to claim 1, wherein a filter size of said second filter processing means is smaller than a filter size of said first filter processing means. 6. A part of a filter coefficient of the second filter processing means is a coefficient obtained by multiplying a part of a filter coefficient of the first filter processing means by n times (n is an integer other than 0). The image processing apparatus according to claim 1 or 5, wherein